Although deaths from cervical cancer in the U.S. have declined dramatically since the implementation of Pap smear screening programs, the prevalence of pre-invasive cervical lesions remains very high. Notably, it is estimated that as few as 12% of high-grade pre-invasive lesions will progress to carcinoma if left untreated, yet all are treated with ablative therapy. Since morphological assessment alone does not allow distinction of those lesions likely to progress from those that will regress or simply persist, the identification of molecular markers that can aid in this distinction and enhance diagnosis and management of pre-invasive cervical lesions would represent a significant clinical advance. Moreover, identification of such markers may lead to an enhanced understanding of the molecular mechanisms that underlie the transition of pre-invasive, clinically harmless lesions, to potentially lethal invasive carcinomas. Although a number of frequent genetic changes in invasive cervical carcinomas have been identified, relatively little is known about the specific molecular alterations that allow pre-invasive cells to acquire the ability to invade the cervical stroma and other features of frankly malignant cells. These alterations, as well as the type of human papillomavirus (HPV) in the lesion and the host's immune response to the HPV infection, likely influence which lesions will progress to invasive cancer. In this application we propose studies aimed at identifying molecular markers associated with, and perhaps underlying, the invasive phenotype. The following Specific Aims are proposed: 1) To employ Affymetrix oligonucleotide microarrays to identify genes differentially expressed in high grade squamous intraepithelial lesions (HSILs) versus invasive cervical carcinomas. Both primary tissues and cell lines will be studied. 2) To determine whether altered expression of selected candidate genes identified in Aim 1 may serve as useful markers for identifying HSILs with increased likelihood of progression to invasive carcinoma. 3) To test selected candidate genes associated with the invasive phenotype for the ability to confer malignant (invasive) properties to HPV-immortalized keratinocytes or cells derived from human squamous intraepithelial lesions.